karpathy

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

rverchere

## Kube State Metrics to get VPA metrics
# Old vs New metrics:
# - kube_verticalpodautoscaler_labels -> kube_customresource_verticalpodautoscaler_labels
# - kube_verticalpodautoscaler_status_recommendation_containerrecommendations_target -> kube_customresource_verticalpodautoscaler_status_recommendation_containerrecommendations_target
# - kube_verticalpodautoscaler_status_recommendation_containerrecommendations_lowerbound -> kube_customresource_verticalpodautoscaler_status_recommendation_containerrecommendations_lowerbound
# - kube_verticalpodautoscaler_status_recommendation_containerrecommendations_upperbound -> kube_customresource_verticalpodautoscaler_status_recommendation_containerrecommendations_upperbound
# - kube_verticalpodautoscaler_status_recommendation_containerrecommendations_uncappedtarget -> kube_customresource_verticalpodautoscaler_status_recommendation_containerrecommendations_uncappedtarget
# Resources:
# - https://github.com/kubernetes/kube-state-metrics/blob/main/docs/customresourcestate-me

m-radzikowski

#!/usr/bin/env bash

set -Eeuo pipefail
trap cleanup SIGINT SIGTERM ERR EXIT

script_dir=$(cd "$(dirname "${BASH_SOURCE[0]}")" &>/dev/null && pwd -P)

usage() {
  cat <<EOF
Usage: $(basename "${BASH_SOURCE[0]}") [-h] [-v] [-f] -p param_value arg1 [arg2...]

akhenakh

        readinessProbe:
          exec:
            command: ["/root/grpc_health_probe", "-addr=:6666"]
          initialDelaySeconds: 1
        livenessProbe:
          exec:
            command: ["/root/grpc_health_probe", "-addr=:6666"]
          initialDelaySeconds: 2
        imagePullPolicy: IfNotPresent

zoidyzoidzoid

#!/usr/bin/env bash
set -e
for full_name in $(kubectl get all -o name | grep -vE replicaset\|pod); do
  name="$(echo "$full_name" | sed -e "s/.*\///g")";
  type="$(echo "$full_name" | sed -e "s/\/.*//g")";
  case "${type}" in
    deployment*)
      type=Deployment
      ;;
    "service")

pydevops

• 0.1. References
• 0.2. Other cheatsheets
• 0.3. Manage multiple gcloud config configurations
  • 0.3.1. Switch gcloud context with gcloud config
• 0.4. Credentials
• 0.5. info
• 0.6. projects
• 0.7. zones & regions
• 0.8. Organization
• 0.9. billing

bwhaley

- hosts: "*"
  become: True
  serial: 1
  tasks:
    - name: Initiate graceful shutdown
      shell: "nodetool {{item}} && sleep 5"
      with_items:
        - disablethrift
        - disablebinary
        - disablegossip

tolitius

worker_processes  1;
error_log logs/error.log;
events {
    worker_connections 1024;
}
http {
  init_worker_by_lua_block {
    redis = require("resty.redis")
  }
  server {

dbaston

FROM phusion/baseimage
 
RUN apt-get update && apt-get install -y \
    build-essential \
    git \
    bison \
    flex \
    zlib1g-dev \
    autoconf \
    libtool \

manelclos

#!/usr/bin/env python
import optparse
import sys
import os

os.environ['S3_USE_SIGV4'] = 'True'

from boto.s3.connection import S3Connection

def sign(bucket, path, access_key, secret_key, https, expiry, host=None):